CN112949503B - Site monitoring management method and system for ice and snow sports - Google Patents
Site monitoring management method and system for ice and snow sports Download PDFInfo
- Publication number
- CN112949503B CN112949503B CN202110243764.3A CN202110243764A CN112949503B CN 112949503 B CN112949503 B CN 112949503B CN 202110243764 A CN202110243764 A CN 202110243764A CN 112949503 B CN112949503 B CN 112949503B
- Authority
- CN
- China
- Prior art keywords
- depth
- potential safety
- target
- safety hazard
- video
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 91
- 238000007726 management method Methods 0.000 title abstract description 33
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000013145 classification model Methods 0.000 claims description 25
- 238000012806 monitoring device Methods 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 14
- 230000003287 optical effect Effects 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 7
- 231100001267 hazard identification Toxicity 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 3
- 230000006378 damage Effects 0.000 claims description 2
- 238000005457 optimization Methods 0.000 claims description 2
- 239000002245 particle Substances 0.000 claims description 2
- 210000000988 bone and bone Anatomy 0.000 description 4
- 230000001133 acceleration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/0202—Child monitoring systems using a transmitter-receiver system carried by the parent and the child
- G08B21/0205—Specific application combined with child monitoring using a transmitter-receiver system
- G08B21/0208—Combination with audio or video communication, e.g. combination with "baby phone" function
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/44—Event detection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- Child & Adolescent Psychology (AREA)
- Theoretical Computer Science (AREA)
- Psychiatry (AREA)
- Human Computer Interaction (AREA)
- Social Psychology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Alarm Systems (AREA)
- Emergency Alarm Devices (AREA)
Abstract
The application discloses a field monitoring and management method and system for ice and snow sports. The method comprises the steps that a depth camera collects a first depth video for preliminary analysis, and if a potential safety hazard target is determined to exist, a safety instruction is sent to a management server; the method comprises the steps that a management server sends query instructions to all depth cameras in a radius area with a current depth camera as a center point; each depth camera searches for a motion track of a potential safety hazard target and sends a second depth video to a management server; and the management server determines the severity of the target potential safety hazard according to the first depth video and the second depth video and selectively sends an alarm instruction to the terminal equipment of the manager. By adopting the technical scheme, the depth camera is used for preliminarily checking the potential safety hazard, and the management server is used for further confirming the potential safety hazard, so that the image data are distributed in different devices for processing, the data processing speed is increased, and the safety of monitoring the ice and snow field is improved.
Description
Technical Field
The application relates to the technical field of communication, in particular to a site monitoring and management method and system for ice and snow sports.
Background
Snow and ice movement, broadly refers to various movements performed on ice and snow, such as: skiing, curling, etc. With the successful declaration of Olympic games in winter in China, the ice and snow sports become more and more full-name sports. However, the existing ice and snow sports field cannot be timely rescued due to the fact that the area is too large to manage, and therefore a method capable of monitoring safety accidents in the field in real time and timely rescuring in real time is urgently needed to ensure the safety of sports or entertainment personnel.
Disclosure of Invention
The application provides a field monitoring method for ice and snow sports, which comprises the following steps:
the method comprises the steps that a field monitoring device collects a depth video of an ice and snow sports field, determines a target person with potential safety hazards from the depth video, and sends a safety instruction to a management server;
after receiving the safety instruction, the management server takes the current depth camera for acquiring the target character as a central point, acquires all site monitoring equipment within a preset radius, and sends a target character query instruction to all site monitoring equipment within the radius area;
each site monitoring device receiving the command of inquiring the target person inquires a video before the potential safety hazard occurs, determines a previous site monitoring device for shooting the previous motion track of the target person from the video, and sends the video to the management server;
the management server determines the motion state of a target character when the potential safety hazard occurs according to the videos shot by the last site monitoring device and the current site monitoring device, inputs the motion state of the target character and the physical quality of the target character into a pre-created model, determines the severity of the potential safety hazard of the target character, and selectively sends an alarm instruction to the terminal device of a manager according to the severity of the potential safety hazard of the target character.
The field monitoring method for the ice and snow sports as described above, wherein the target person with the potential safety hazard is determined from the depth video, specifically includes the following sub-steps:
the site monitoring equipment extracts a target person depth image from the depth video and extracts a target person skeleton coordinate from an image frame;
the site monitoring equipment preliminarily identifies the initial potential safety hazard degree of the target person according to the target human body proportion and the target person skeleton coordinate in the human body depth image, and sends the human body depth image and the shooting time to the management server;
the management server takes the depth camera for acquiring the falling target as a central point, acquires images of all depth cameras within a preset radius before an accident occurs, and sends instructions to all the depth cameras;
each depth camera extracts an image frame with a target person from a video shot by the depth camera, the depth camera extracted with the target person is used as a previous depth camera, and the accurate potential safety hazard degree of the target person is identified based on the motion trend of the target person in the image frames of the current depth camera and the previous depth camera.
The above field monitoring method for the ice and snow sports, wherein the extracting of the depth image of the target person from the depth video specifically includes the following substeps:
graying the human body depth image to obtain a relatively soft gray image;
carrying out difference processing on the current gray level image and a preset background image and carrying out mathematical morphology processing to obtain a motion framework of the moving target:
the method comprises the steps of storing an original image of an unmanned object in an ice and snow field in a depth camera, using the original image as a preset background image, carrying out filtering pretreatment on a gray image, carrying out background difference treatment on the gray image and the preset background image to extract a target area, and increasing the contrast of the target area.
The method for monitoring the field for the ice and snow sports comprises the steps of defining all possible gesture sets in advance according to the height-width ratio of a rectangular frame where the position of a human body is located, determining the area combination according to each gesture, marking the rectangular frame where the human body is located from a human body depth image, extracting the height and width of a target human body from the rectangular frame, and preliminarily identifying the initial potential safety hazard degree of the target person according to the height-width ratio.
According to the field monitoring method for the ice and snow sports, the skeleton coordinates of the target person and the height and width of the rectangular frame of the human body are obtained from the depth video shot by the last depth camera, the human body pose is determined, the pose change of the target person in a short time is determined according to the comparison between the human body pose of the camera and the human body pose shot during falling, and the accurate potential safety hazard degree of the target person is determined.
The field monitoring method for the ice and snow sports is characterized in that the model is obtained by training a human body motion state, a human body quality and an impact force which can be born by the human body in a large quantity.
The method for monitoring the field for the ice and snow sports comprises the steps of obtaining images of all cameras within a preset radius before an accident occurs by taking the above one-depth camera as a central point after obtaining the last motion track of a target person, determining another last motion track of the target person shot from the images, and repeatedly searching the track.
The application also provides a site monitoring system for the ice and snow sports, which is characterized by comprising site monitoring equipment, a management server and manager terminal equipment, wherein the system executes any one of the site monitoring methods for the ice and snow sports.
The field monitoring system for the ice and snow sports as described above, wherein the field monitoring device includes a plurality of side-by-side monitoring piles symmetrically installed at both ends of the field, a cableway is installed between each opposing set of monitoring piles, a monitoring cable car is slidably connected on the cableway, the monitoring cable car slides between the set of monitoring piles at a preset speed, and a depth camera is installed at the bottom of the monitoring cable car.
As above, the site monitoring system for the ice and snow sports, wherein the depth camera arranged at the bottom of the monitoring cable car is used for shooting the video of the ice and snow sports site, the optical cables for data are connected between all monitoring piles and on the cableway, the depth camera is used for preliminarily identifying the shot site video, then the video of the ice and snow sports site shot by the depth camera is transmitted to the management server through the optical cables for analyzing and monitoring the ice and snow sports site, and whether an alarm instruction is sent to the terminal equipment of a manager or not is determined according to the monitoring result.
The beneficial effect that this application realized is as follows: by adopting the technical scheme, the site monitoring equipment is used for preliminarily checking the potential safety hazard, and the management server is used for further confirming the potential safety hazard, so that the image data can be distributed in different equipment for processing, the data processing speed is increased, and the monitoring safety of the ice and snow site is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings can be obtained by those skilled in the art according to the drawings.
Fig. 1 is a flowchart of a field monitoring method for an ice and snow sports according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example one
The embodiment of the application provides a site monitoring method for ice and snow sports, which is applied to a site monitoring system, wherein the site monitoring system comprises site monitoring equipment, a management server and manager terminal equipment; the field monitoring equipment comprises a plurality of side-by-side monitoring piles symmetrically installed at two ends of a field, a cableway is installed between every two opposite groups of monitoring piles, a monitoring cable car is connected on the cableway in a sliding mode and slides between the groups of monitoring piles at a preset speed, a depth camera arranged at the bottom of the monitoring cable car is used for shooting a video of the ice and snow sports field, optical cables of data are connected between all the monitoring piles and the cableway, and the video of the ice and snow sports field shot by the depth camera is transmitted to a server through the optical cables to analyze and monitor the ice and snow sports field; and determining whether to send an alarm instruction to the manager terminal equipment or not according to the monitoring result.
Specifically, a depth camera acquires a first depth video, performs preliminary analysis on the first depth video, and sends a safety instruction to a management server if a potential safety hazard target is determined to exist; the management server takes the depth camera as a central point, acquires all depth cameras within a preset radius, and sends query instructions to all depth cameras within the radius area; each depth camera receiving the query instruction searches a motion track of the potential safety hazard target from the depth video shot by the depth camera, and sends a second depth video shot by the depth camera which finds the potential safety hazard target to the management server; and the management server determines the severity of the target potential safety hazard according to the first depth video and the second depth video, and selectively sends an alarm instruction to the terminal equipment of the manager according to the severity of the potential safety hazard of the target person.
As shown in fig. 1, the method for monitoring a field for ice and snow sports specifically includes the following steps:
step 110, collecting a depth video of an ice and snow sports field by field monitoring equipment, determining a target person with potential safety hazards from the depth video, and sending a safety instruction to a management server;
specifically, in the field monitoring system, a depth video in a field is shot in real time by a depth camera installed at the bottom of a monitoring cable car on a cableway, and the video shot in a preset time length and shooting time are recorded for searching a target person from the shot video subsequently and determining the state of the target person before an accident occurs;
the method for determining the target person with the potential safety hazard from the depth video specifically comprises the following substeps:
111, extracting a target person depth image from the depth video and extracting a target person skeleton coordinate from an image frame by using site monitoring equipment;
creating a skeleton event in a depth camera, opening a skeleton tracking function, and collecting a human body depth image and an RGB (red, green and blue) image containing skeleton data; specifically, the human body depth image is grayed firstly: in the color model RGB, when R ═ G ═ B, then the color (R, G, B) is represented as a black and white color; the process of making the R, G, B component values of the color equal is graying; because the value ranges of R, G and B are 0-255, the gray level is only 256 levels, and the gray image is calculated by adopting the following formula: r ═ G ═ B ═ R + G + B)/3, and a relatively soft grayscale image was obtained. Then, carrying out difference processing on the current gray level image and a preset background image, and carrying out mathematical morphology processing to obtain a motion framework of the moving target: the method comprises the steps of storing an original image of an unmanned object in an ice and snow field in a depth camera, using the original image as a preset background image, carrying out filtering pretreatment on a gray image, carrying out background difference treatment on the gray image and the preset background image to extract a target area, and increasing the contrast of the target area.
Step 112, the site monitoring equipment preliminarily identifies the initial potential safety hazard degree of the target person according to the target human body proportion and the target person skeleton coordinate in the human body depth image, and sends the human body depth image and the shooting time to a management server;
in the embodiment of the application, all possible gesture sets are defined in advance according to the height-width ratio of a rectangular frame where a human body is located, a regional combination of the gesture sets is determined according to each gesture, the rectangular frame where the human body is located is marked out of a human body depth image, then the height and width of a target human body are extracted from the rectangular frame, the normal ratio of the height to the width of the normal human body is 3-5 when the normal human body stands, and if the height-width ratio of the target human body of the current human body depth image is calculated to be 0.2-1, the possibility that the target person falls down is preliminarily identified; and the possibility of falling of the target person can be preliminarily identified by comparing the bone coordinates of the target person with the bone coordinates of the standard human body.
Step 113, the management server takes the depth camera for obtaining the falling target as a central point, obtains images of all depth cameras within a preset radius before an accident occurs, and sends instructions to all depth cameras;
the method comprises the steps that positions of all depth cameras are stored in a management server, unique identification is marked for each depth camera, when the depth cameras within a preset radius need to be determined, an instruction is sent to the corresponding depth cameras, and the instruction comprises the unique identification of the depth cameras, the shooting time of the current depth cameras and a target person image frame.
Step 114, each depth camera extracts an image frame with a target person from a video shot by the depth camera, the depth camera extracted with the target person is used as a previous depth camera, and the accurate potential safety hazard degree of the target person is identified based on the motion trend of the target person in the image frames of the current depth camera and the previous depth camera;
after receiving the instruction of the management server, each depth camera searches for a video (generally within 10 s) shot by itself before the shooting time of the current depth camera, and searches whether a target person exists in the video, wherein the depth camera with the target person is a motion state shot at the last moment of the target person.
Because the depth camera preliminarily judges whether the human body falls down or not, the possibility that the target person independently squats down may exist, and whether the target person falls down or not needs to be further accurately identified according to the pose change of the target human body within a certain short time period (for example, if the first 3s shooting is a standing pose, and if the current shooting is a falling pose, the target person is determined to fall down), a depth video before an accident is taken by all cameras within a circle is taken as a central point, the cameras shooting the target person are found out from the depth video shot by the cameras, the bone coordinates of the target person and the height and width of a rectangular frame of the human body are obtained, the pose of the human body is determined, and the pose change of the target person within a short time period is determined according to the comparison between the pose of the cameras and the pose of the human body shot during the falling down, and determining that the target person has a falling event.
Step 120, after receiving the safety instruction, the management server takes the current depth camera for acquiring the target person as a central point, acquires all site monitoring devices within a preset radius, and sends a target person query instruction to all site monitoring devices within the radius area;
step 130, inquiring videos before potential safety hazards occur by each site monitoring device receiving the command of inquiring the target person, determining a last site monitoring device for shooting a last motion track of the target person, and sending the videos to a management server;
step 140, the management server determines the motion state of the target person when the potential safety hazard occurs according to the videos shot by the previous site monitoring device and the current site monitoring device, inputs the motion state of the target person and the physical quality of the target person into a pre-created model, determines the severity of the potential safety hazard of the target person, and selectively sends an alarm instruction to the terminal device of the manager according to the severity of the potential safety hazard of the target person;
in the embodiment of the application, determining the severity of the potential safety hazard of the target person specifically comprises the following substeps:
141, acquiring different motion conditions of different body qualities and bearable impact force to obtain feature vectors, inputting the feature vectors into a classification model, training the classification model to obtain different sub-classification models, classifying a feature vector set by using each sub-classification model respectively, and estimating a set of weights of each sub-classification model according to a classification result;
specifically, a large amount of data of different physical qualities such as different sexes, heights and widths in historical falling events are collected, the movement conditions of people with different physical qualities during falling, such as the speed and the acceleration during falling, the impact force which can be borne by the people, namely the physical injuries of different degrees which can be caused by different force ranges are collected, and the data are combined into a feature vector set A { (x) 1 ,y 1 ,z 1 ,t 1 ),(x 2 ,y 2 ,z 2 ,t 2 ),(x 3 ,y 3 ,z 3 ,t 3 )…(x n ,y n ,z n ,t 4 ) In which x 1 ,x 2 …x n For sex characteristics (1 and 2 only, 1 for male and 2 for female), y 1 ,y 2 …y n To correspond to aspect ratio features, z 1 ,z 2 …z n Corresponding to the speed characteristic when the person falls down, t is the impact force which can be born correspondingly, and n is the number of collected samples; inputting the feature vector set into the classification model, and training a sub-classification model f by using the feature vector set t (t); reuse sub-classification model f t (t) classifying the feature vector set to obtain a classification result, and adopting a formula according to the classification result
Step 142, searching an optimal value corresponding to each weight in the weight set, and determining the severity of the potential safety hazard through the combination of each sub-classification model and the optimal value of the weight corresponding to the sub-classification model to obtain a potential safety hazard identification model;
calculating each sub-classification model f by particle swarm optimization algorithm t Set of weights of (t) { μ } 1 ，μ 2 ，μ 3 ......μ T In the method, each weight corresponds to an optimal value; by each sub-classification model { f 1 (t)、f 2 (t)、f 3 (t)......f T (t) } and its corresponding optimal value of weight [ mu ] 1 ，μ 2 ，μ 3 ......μ T And (6) determining the severity of the potential safety hazard in a combined manner to obtain a potential safety hazard identification model.
Step 243, determining the physical quality of the target person from the video shot by the previous depth camera, calculating the movement state of the target person when the target person falls down from the video shot by the current depth camera and the video shot by the previous depth camera, inputting the movement state and the physical quality of the target person into a potential safety hazard identification model, and outputting the severity of the potential safety hazard;
preferably, the height and the width of the target person can be determined from the video of the last depth camera, the gender of the target person can be determined according to bone data in the depth video, the video shot by the current depth camera and the last depth camera is used for calculating the falling speed of the target person, the height, the width and the gender of the target person and the falling speed of the target person are input into a potential safety hazard recognition model, the falling impact force of the target person is calculated, if the falling impact force of the target person is higher than the maximum impact force which can be borne by the human body, the high-numerical potential safety hazard severity is output, an alarm instruction is sent to a manager after the high-numerical potential safety hazard severity is output, and the manager is instructed to rescue in time.
Example two
The second embodiment of the application provides a site monitoring system for ice and snow sports, which comprises site monitoring equipment, a management server and manager terminal equipment, wherein the system executes the site monitoring method for ice and snow sports in the first embodiment.
The site monitoring equipment comprises a plurality of side-by-side monitoring piles symmetrically installed at two ends of a site, a cableway is installed between every two opposite groups of monitoring piles, a monitoring cable car is connected on the cableway in a sliding mode and slides at a preset speed between the groups of monitoring piles, and a depth camera is installed at the bottom of the monitoring cable car.
The method comprises the steps that a depth camera arranged at the bottom of a monitoring cable car is used for shooting a video of the ice and snow sports field, optical cables of data are connected between all monitoring piles and on a cableway, the depth camera is used for preliminarily recognizing the shot field video, then the ice and snow sports field video shot by the depth camera is transmitted to a management server through the optical cables to be analyzed and monitored in the ice and snow sports field, and whether an alarm instruction is sent to a terminal device of a manager or not is determined according to a monitoring result.
The above-mentioned embodiments are only specific embodiments of the present application, and are used for illustrating the technical solutions of the present application, but not limiting the same, and the scope of the present application is not limited thereto, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the present disclosure, which should be construed in light of the above teachings. Are intended to be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A field monitoring method for ice and snow sports, comprising:
the method comprises the steps that a depth camera collects a first depth video, preliminary analysis is conducted on the first depth video, and if a potential safety hazard target is determined to exist, a safety instruction is sent to a management server;
the management server takes the depth camera as a central point, acquires all depth cameras within a preset radius, and sends query instructions to all depth cameras within the radius area;
each depth camera receiving the query instruction searches a motion track of the potential safety hazard target from the depth video shot by the depth camera, and sends a second depth video shot by the depth camera which finds the potential safety hazard target to the management server;
the management server determines the severity of the target potential safety hazard according to the first depth video and the second depth video, and selectively sends an alarm instruction to the terminal equipment of the manager according to the severity of the potential safety hazard of the target person;
determining the severity of the potential safety hazard of the target person, specifically comprising the following substeps:
acquiring different motion conditions of different body qualities and bearable impact force to obtain characteristic vectors, inputting the characteristic vectors into a classification model, training the classification model to obtain different sub-classification models, classifying a characteristic vector set by using each sub-classification model respectively, and estimating a set of weights of each sub-classification model according to a classification result;
in particular, miningThe method comprises the steps of collecting a large amount of data of different physical qualities of different sexes and aspect ratios in historical falling events, collecting the motion conditions of people with different physical qualities when falling, and the impact force born by the people, namely the physical injuries of different degrees possibly caused by different force ranges, and forming a characteristic vector set A { (x) 1 ,y 1 ,z 1 ,t 1 ),(x 2 ,y 2 ,z 2 ,t 2 ),(x 3 ,y 3 ,z 3 ,t 3 )…(x n ,y n ,z n ,t 4 ) In which x 1 ,x 2 …x n As a sex attribute, y 1 ,y 2 …y n To correspond to aspect ratio features, z 1 ,z 2 …z n Corresponding to the speed characteristic when the person falls down, t is the impact force which can be born correspondingly, and n is the number of collected samples; inputting the feature vector set into the classification model, and training a sub-classification model f by using the feature vector set t (t); reuse sub-classification model f t (t) classifying the feature vector set to obtain a classification result, and adopting a formula according to the classification result
Estimating a set of weights { μ ] of a subcategory model 1 ，μ 2 ，μ 3 ......μ T }; wherein argmin is
A set of μ with a minimum value;
searching an optimal value corresponding to each weight in the weight set, and determining the severity of the potential safety hazard through the combination of each sub-classification model and the optimal value of the weight corresponding to the sub-classification model to obtain a potential safety hazard identification model;
calculating each sub-classification model f by particle swarm optimization algorithm t Set of weights of (t) { μ } 1 ，μ 2 ，μ 3 ......μ T In the method, each weight corresponds to an optimal value; by each sub-classification model { f 1 (t)、f 2 (t)、f 3 (t)……f T (t) } and its corresponding optimal value of weight [ mu ] 1 ，μ 2 ，μ 3 ……μ T Determining the severity of the potential safety hazard in a combined mode to obtain a potential safety hazard identification model;
the method comprises the steps of determining the height and the width of a target character from a video of a previous depth camera, determining the gender of the target character according to skeleton data in the depth video, calculating the falling speed of the target character from the video shot by the current depth camera and the previous depth camera, inputting the height, the width and the gender of the target character and the falling speed of the target character into a potential safety hazard recognition model, calculating the falling impact force of the target character, outputting the high-value potential safety hazard severity if the falling impact force of the target character is higher than the maximum impact force which can be borne by a human body, and sending an alarm instruction to a manager after outputting the high-value potential safety hazard severity to instruct the manager to rescue in time.
2. The field monitoring method for snowy and icy sports of claim 1,
the method comprises the steps that a field monitoring device collects a depth video of an ice and snow sports field, determines a target person with potential safety hazards from the depth video, and sends a safety instruction to a management server;
after receiving the safety instruction, the management server takes the current depth camera for acquiring the target character as a central point, acquires all site monitoring equipment within a preset radius, and sends a target character query instruction to all site monitoring equipment within the radius area;
each site monitoring device receiving the command of inquiring the target person inquires a video before the potential safety hazard occurs, determines a previous site monitoring device for shooting the previous motion track of the target person from the video, and sends the video to the management server;
the management server determines the motion state of a target character when the potential safety hazard occurs according to the videos shot by the last site monitoring device and the current site monitoring device, inputs the motion state of the target character and the physical quality of the target character into a pre-created model, determines the severity of the potential safety hazard of the target character, and selectively sends an alarm instruction to the terminal device of a manager according to the severity of the potential safety hazard of the target character.
3. A site monitoring method for ice and snow sports as claimed in claim 2, wherein the step of determining the target person with potential safety hazard from the depth video comprises the following sub-steps:
the site monitoring equipment extracts a target person depth image from the depth video and extracts a target person skeleton coordinate from an image frame;
the site monitoring equipment preliminarily identifies the initial potential safety hazard degree of the target person according to the target human body proportion and the target person skeleton coordinate in the human body depth image, and sends the human body depth image and the shooting time to the management server;
the management server takes the depth camera for acquiring the falling target as a central point, acquires images of all depth cameras within a preset radius before an accident occurs, and sends instructions to all the depth cameras;
each depth camera extracts an image frame with a target person from a video shot by the depth camera, the depth camera extracted with the target person is used as a previous depth camera, and the accurate potential safety hazard degree of the target person is identified based on the motion trend of the target person in the image frames of the current depth camera and the previous depth camera.
4. A site monitoring method for ice and snow sports as claimed in claim 3, wherein extracting the depth image of the target person from the depth video comprises the following sub-steps:
graying the human body depth image to obtain a relatively soft gray image;
carrying out difference processing on the current gray level image and a preset background image and carrying out mathematical morphology processing to obtain a motion framework of the moving target:
the method comprises the steps of storing an original image of an unmanned object in an ice and snow field in a depth camera, using the original image as a preset background image, carrying out filtering pretreatment on a gray image, carrying out background difference treatment on the gray image and the preset background image to extract a target area, and increasing the contrast of the target area.
5. The field monitoring method for ice and snow sports as claimed in claim 3, wherein all possible pose sets are defined in advance according to the ratio of the height and width of a rectangular frame where the human body is located, the regional combination of the pose sets is determined according to each pose, the rectangular frame where the human body is located is marked out from the human body depth image, then the height and width of the target human body are extracted from the rectangular frame, and the initial potential safety hazard degree of the target person is preliminarily identified according to the height and width ratio.
6. The field monitoring method for the ice and snow sports of claim 4, wherein the skeletal coordinates of the target person and the height and width of the rectangular frame of the human body are obtained from the depth video shot by the previous depth camera, the pose of the human body is determined, the pose change of the target person in a short time is determined according to the comparison between the pose of the human body of the camera and the pose of the human body shot when the target person falls, and the accurate potential safety hazard degree of the target person is determined.
7. The field monitoring method for ice and snow sports of claim 1, further comprising acquiring images of all cameras within a predetermined radius before an accident occurs by using the above-depth camera as a central point after acquiring the previous motion trajectory of the target person, determining therefrom a further previous motion trajectory of the target person photographed, and repeating the trajectory search.
8. A site monitoring system for an ice and snow sports, characterized in that the system includes site monitoring devices, a management server and manager terminal devices, and the system performs the site monitoring method for ice and snow sports as set forth in any one of claims 1 to 7.
9. A field monitoring system for ice and snow sports as claimed in claim 8, wherein the field monitoring apparatus comprises a plurality of side by side monitoring piles symmetrically installed at both ends of the field, a cableway is installed between each opposing set of monitoring piles, a monitoring trolley is slidably connected on the cableway, the monitoring trolley slides between the set of monitoring piles at a preset speed, and a depth camera is installed at the bottom of the monitoring trolley.
10. A site monitoring system for an icy and snowy sports according to claim 9, wherein the video of the icy and snowy sports site is shot by a depth camera installed at the bottom of the monitoring cable car, optical cables for data are connected between all monitoring piles and on the cableway, the depth camera performs preliminary identification on the shot site video, then the video of the icy and snowy sports site shot by the depth camera is transmitted to the management server through the optical cables for analysis and monitoring of the icy and snowy sports site, and whether to send an alarm instruction to the terminal device of the manager is decided according to the monitoring result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110243764.3A CN112949503B (en) | 2021-03-05 | 2021-03-05 | Site monitoring management method and system for ice and snow sports |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110243764.3A CN112949503B (en) | 2021-03-05 | 2021-03-05 | Site monitoring management method and system for ice and snow sports |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112949503A CN112949503A (en) | 2021-06-11 |
| CN112949503B true CN112949503B (en) | 2022-08-09 |
Family
ID=76247816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110243764.3A Active CN112949503B (en) | 2021-03-05 | 2021-03-05 | Site monitoring management method and system for ice and snow sports |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112949503B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056035A (en) * | 2016-04-06 | 2016-10-26 | 南京华捷艾米软件科技有限公司 | Motion-sensing technology based kindergarten intelligent monitoring method |
| WO2017133009A1 (en) * | 2016-02-04 | 2017-08-10 | 广州新节奏智能科技有限公司 | Method for positioning human joint using depth image of convolutional neural network |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AT500170A3 (en) * | 2002-12-18 | 2008-09-15 | Baumgartner Leo | CABLE CAR FOR THE CARRIAGE OF CAMERAS AND, IF APPLICABLE, A CAMERA NAMEL |
| US9235765B2 (en) * | 2010-08-26 | 2016-01-12 | Blast Motion Inc. | Video and motion event integration system |
| US9396385B2 (en) * | 2010-08-26 | 2016-07-19 | Blast Motion Inc. | Integrated sensor and video motion analysis method |
| BE1021528B1 (en) * | 2013-02-01 | 2015-12-08 | Familyeye Bvba | FALL DETECTION SYSTEM AND METHOD FOR DETECTING A FALL OF A MONITORED PERSON |
| US20140276238A1 (en) * | 2013-03-15 | 2014-09-18 | Ivan Osorio | Method, system and apparatus for fall detection |
| US20180137363A1 (en) * | 2015-04-03 | 2018-05-17 | Mas-Tech S.R.L. | System for the automated analisys of a sporting match |
| US20170281054A1 (en) * | 2016-03-31 | 2017-10-05 | Zoll Medical Corporation | Systems and methods of tracking patient movement |
| US20170309152A1 (en) * | 2016-04-20 | 2017-10-26 | Ulysses C. Dinkins | Smart safety apparatus, system and method |
| CN109460702B (en) * | 2018-09-14 | 2022-02-15 | 华南理工大学 | Passenger abnormal behavior identification method based on human body skeleton sequence |
| CN109871775A (en) * | 2019-01-22 | 2019-06-11 | 北京影谱科技股份有限公司 | A kind of the ice rink monitoring method and device of Behavior-based control detection |
| RU2718223C1 (en) * | 2019-07-18 | 2020-03-31 | ООО "Ай Ти Ви групп" | System and method for determining potentially dangerous situations based on video data |
| CN111767888A (en) * | 2020-07-08 | 2020-10-13 | 北京澎思科技有限公司 | Object state detection method, computer device, storage medium, and electronic device |
-
2021
- 2021-03-05 CN CN202110243764.3A patent/CN112949503B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017133009A1 (en) * | 2016-02-04 | 2017-08-10 | 广州新节奏智能科技有限公司 | Method for positioning human joint using depth image of convolutional neural network |
| CN106056035A (en) * | 2016-04-06 | 2016-10-26 | 南京华捷艾米软件科技有限公司 | Motion-sensing technology based kindergarten intelligent monitoring method |
Non-Patent Citations (2)
| Title |
|---|
| "Evaluation of a threshold-based tri-axial accelerometer fall detection algorithm";A.K.Bourke等;《Gait & Posture》;20070731;第26卷(第2期);全文 * |
| "基于分类学习的住院老年人跌倒行为检测研究";姜珊;《中国优秀博硕士学位论文全文数据库(硕士) 医学卫生科技辑》;20200815(第08期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112949503A (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109522793B (en) | Method for detecting and identifying abnormal behaviors of multiple persons based on machine vision | |
| CN107911663A (en) | A kind of elevator passenger hazardous act intelligent recognition early warning system based on Computer Vision Detection | |
| CN110765964A (en) | Method for detecting abnormal behaviors in elevator car based on computer vision | |
| CN111401311A (en) | High-altitude parabolic recognition method based on image detection | |
| CN110309718A (en) | A kind of electric network operation personnel safety cap wearing detection method | |
| JP4956273B2 (en) | Throwing ball type discriminating device, discriminator generating device, throwing ball type discriminating program and discriminator generating program | |
| CN109829382B (en) | Abnormal target early warning tracking system and method based on intelligent behavior characteristic analysis | |
| KR20150021526A (en) | Self learning face recognition using depth based tracking for database generation and update | |
| CN113963315A (en) | Real-time video multi-user behavior recognition method and system in complex scene | |
| CN113191699A (en) | Power distribution construction site safety supervision method | |
| CN111401310B (en) | Kitchen sanitation safety supervision and management method based on artificial intelligence | |
| CN111047874A (en) | Intelligent traffic violation management method and related product | |
| KR20210062256A (en) | Method, program and system to judge abnormal behavior based on behavior sequence | |
| CN107330918B (en) | Football video player tracking method based on online multi-instance learning | |
| CN114360209B (en) | Video behavior recognition security system based on deep learning | |
| CN112949503B (en) | Site monitoring management method and system for ice and snow sports | |
| CN106934339B (en) | Target tracking and tracking target identification feature extraction method and device | |
| CN112885014A (en) | Early warning method, device, system and computer readable storage medium | |
| CN112464765A (en) | Safety helmet detection algorithm based on single-pixel characteristic amplification and application thereof | |
| CN115880620B (en) | Personnel counting method applied to cart early warning system | |
| CN116843726A (en) | Pedestrian track tracking method and device, electronic equipment and storage medium | |
| CN115439925A (en) | Action scoring method, terminal and computer readable storage medium | |
| CN111582278A (en) | Portrait segmentation method and device and electronic equipment | |
| US20210287051A1 (en) | Methods and systems for recognizing object using machine learning model | |
| CN111062294B (en) | Passenger flow queuing time detection method, device and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240516 Address after: Room 111-007, Floor 1, Building 2, Science and Technology Innovation Headquarters, Shenzhen (Harbin) Industrial Park, 288 Zhigu Street, Songbei District, Harbin, Heilongjiang Province, 150028 Patentee after: Harbin Safety Measurement and Control Technology Co.,Ltd. Country or region after: China Address before: 161006, No. 42, culture street, Jianhua District, Heilongjiang, Qigihar Patentee before: QIQIHAR University Country or region before: China |